i just took a look here and found out that right now St. Nicolaus-day will deliver a new NC-Major version to our shoes and boots.
It would be the 3rd major-version which would come out during 2018, with the last ‘new’ one (NC14) having been released sometime in September 2018 - roughly 3 months earlier, only.
After having gone through several severe problems during the upgrade to NC14 (at least following to comments here on the forum, as i myself am still sticking to 13.0.7 right now) I am not sure if a release of a new major version as soon after nc14 would be neccessary. Maybe it would have been better to skip one release? I really don’t know… and so I would be glad to have someone trying to explain the fast release-schedule.
I mean we would need to keep in mind that major releases like that would be really demandive not only to users but to several app-devs as well who would need to apply changes to their apps as well.
plus: what could we learn for the future from the (bumpy?) release of NC14?
and here’s a plea: we have read enough rants about the release of NC14 - so no need to vent off here again, that’s simply unneccessary. I want to really learn from the happenings to improve the improve NC in the future. for everyones benefit.
Let me just raise some points “from the outside”, as I am not related to Nextcloud GmbH
3 releases in 2016
1 releases in 2017
it will have 3 releases in 2018
If you look at it, we do not have enough data but the “strange” year could have been 2017
Nextcloud GmbH has to hit the right balance between having a reliable product and having a product with more functionalities than the competitors. Tough endeavour for everybody
Nextcloud (the firm and the community) is trying to improve the life of the app developers; as far as I understood that is the purpose of the Vue.js components, reusable by the app developers. Of course the improvement will lead to the need of modifications of the apps…
Nextcloud is developing some interesting features. I do agree that “featuritis” is a bad disease for software. Yet I still have the impression that some good ideas about collaboration that @frank illustrated years ago in some speeches still require a lot of features to be developed. The final goal was not yet reached, they are trying to reach it and they know that they cannot be too slow…
I do agree that we need both solid software and satisfied sysadmins and app developers. I hope that Nextcloud will manage to reach this goal
More version means also fewer changes between the versions. So there are probably less problems during an upgrade and also less modifications for app developers. I’d prefer doing 3 painless upgrades per year than doing one with many problems.
It’s the planned date. In the past, this has been the most optimistic case and I would expect the real release date end of December or even in January.
Yes you can have the impression that this upgrade was a bit more problematic than previous ones. So it would be important to make sure all problems are well understood and that they will be fixed for the next release. Not sure if everything was well identified and reported to github.
of course i am aware of that. as i am even one of the guys saying goodbye to some prescheduled release-plans in order to just keep the date rather than having a better tested and prolly better working release some days/weeks/months later. i prefer the latter.
If I may add my 2c. I think if you look at NC holistically, it is a very fast-paced changing product. Some may argue it is playing catch-up with the “evil twins”, some may argue it is trying to set the benchmark. Whatever the case may be, I think 2-3 releases per year is quite the norm for new open source software. Hell, I have been using a lecture capture solution (open source) which in one year nearly did 4 releases. So it is not something new to me that a “new kid on the block” has this many releases.
I understand completely your concern for staying up to date and within support boundaries. I have no concrete answer for your concern, all that I do know is that we should trust the devs when they release the product major versions, and as @tflidd said, maybe make sure github bugs are filed in time. This will create an environment where the devs are aware of “real-world” problems, and can get to a point where they can clearly see if the upgrade schedule is too fierce on their consumers.
Generally it is a good sign that Nextcloud develops that fast, that new features, as well requested by users, are implemented in a reasonable schedule. Jep and of course it is as well about the development race with the direct concurrent to keep or increase marked share.
Of course faster development directly affects stability and support length, so a balance need to be found. To keep or increase stability/reliability the beta and RC phases are done and announced with a special eye on the update process, offering little rewards (NC T-shirts it was mostly) for new found update bugs, as a motivation to cover a higher amount of setups.
Perhaps these testing phases need to be enlarged, announced more widely? I remember the Nextcloud announcement app, which was for official notifications from Nextcloud side, but the last time I had it active, I never received something there? Perhaps this could be more actively used to reach more users that do not regularly have a look into the forum? But maybe this is already the case, didn’t use that app for a while.
But generally, this is the deal with open source business models. Either regularly update the product to stay with current security updates, but having possibly more work and less stability. Or pay for a LTS plan, if you need max security with max reliability and minimum down time of your that important services, usually required for larger companies. This is how Nextcloud GmbH earns money .
Definitely better than a product that stalls in development, can’t catch up with concurrents, thus loses market share and the ability to pay enough developers for reliable dev to stay with professional use.